Candida antifungal drug resistance in sub-Saharan African populations: A systematic review

Introduction

Candida species are known to shift from commensal to opportunistic infectious agents when triggered by factors such as immunosuppression, continuous usage of antibiotics and poor nutrition, leading to increased patient morbidity and mortality^1–3. In severely immunocompromised patients, Candida species can spread through the bloodstream and gastrointestinal tract. This can lead to systemic candidiasis, with a reported mortality rate in developed countries of 38%⁴ and 44%⁵. Candida is currently the 4^th most commonly isolated microorganism in nosocomial bloodstream infections⁶ and has been implicated in >78% of cancerous and precancerous oral lesions⁷.

Various antifungal drugs with different modes of action have been developed over the years. These include polyene antifungals (e.g. nystatin and amphotericin B), which interfere with ergosterol synthesis, thereby causing cell membrane leakage; the imidazoles (e.g. miconazole, clotrimazole, econazole and ketoconazole), which also interfere with ergosterol and other cell membrane sterol synthesis; the echinocandins (e.g. anidulafungin, micafungin and caspofungin), which inhibit β 1–3 glucan synthesis, affecting the fungal cell wall and 5-flucytosine that in turn interferes with fungal RNA and DNA synthesis⁸. The triazoles (including fluconazole, posaconazole, voriconazole and itraconazole) interfere with the synthesis of ergosterol and have been shown to have fewer side effects than some of the other antifungal drug classes⁹.

Resistance to available antifungal therapies is widespread^10,11, probably due to the widespread and repeated use of these drugs¹². Different Candida species have varying resistance patterns, which appear to be geographically determined^13,14. Therefore early recognition of resistance facilitates the selection of an appropriate antifungal drug, with the use of oral antifungals in oropharyngeal candidiasis reserved for cases where there is no response to topical antifungal treatment¹⁵. Resistance pattern surveillance to avoid an even higher number of improperly treated, and therefore resistant fungal infections, is imperative¹⁶. This is a cause for concern in the case of immunocompromised patients, who are at a much higher risk of developing opportunistic complications. Importantly, sub-Saharan Africa is the region most affected by HIV, with approximately 25.8 million infected people in 2014 and accounting for almost 70% of the global number of new HIV infections (http://www.who.int/mediacentre/factsheets/fs360/en/).

Programmes on species prevalence and antifungal surveillance have been successfully developed and introduced in Europe, Asia-Pacific, Latin America and North America^17–19. The gap in antifungal drug resistance surveillance in Africa has been documented²⁰. Surveillance programmes are crucial tools in the transition from empirical antifungal treatment, which often does not work due to the diverse resistance levels seen in different regions, and the presence of species that are intrinsically resistant to certain antifungal drugs. The non-existence of routine diagnostics laboratories in most African countries has meant that many African patients are treated without the knowledge of which species they harbour and without any updated guideline data that could be used as a reference in prescribing antifungals. Possible causes for the lack of Candida surveillance programmes in Africa include lack of funding, the limited number of research collaborations and the existence of conflict areas within the continent. This prompted the need for a review of the current situation in Africa regarding the drug susceptibility profiles of Candida species available from different regions.

Methods

Results and discussion

The study populations included healthy^22,23, HIV-positive^22–35 and cancer patients²³, as well as patients with genitourinary tract infections^36–39, respiratory tract infections^32,39, meningitis³⁹ and candidemia^40,41. Most studies relied on broth microdilution or disk diffusion for antimicrobial susceptibility testing, while one of the publications was a retrospective clinical study based on the patients’ response to antifungal therapy.

This review included seven studies from two regions in South Africa^{22–25,36,40,41}, three studies from different regions in Ethiopia^33–35, three studies from two regions in Cameroon^25–27, three studies from different regions in Nigeria^28,29,37, two studies from the same region in Ivory Coast^30,38, one study from Tanzania³¹, one study from Kenya³² and one study from Ghana³⁹. Due to the paucity of studies and differences in isolation and antifungal susceptibility testing, a meta-analysis could not be conducted.

Non-albicans species, such as C. glabrata and C. krusei are reported to have innate resistance to antifungal drugs. C. krusei resistance has been reported from South Africa^22,25, Cameroon²⁵, Nigeria^28,37, Ghana³⁹, Tanzania³¹ and Ethiopia³⁴. Initially thought to have innate resistance to azoles, C. glabrata resistance has been reported in Cameroon²⁵, Ethiopia³⁴ and Tanzania³¹, while susceptibility has been reported in South Africa^22,25 and Nigeria²⁸. This discrepancy may be explained by the phenotypic similarity between C. glabrata, C. nivariensis and C. bracarensis, which could possibly be confused in the absence of molecular typing methods and show different antifungal profiles⁴². Resistant C. glabrata has increased in patients presenting with candidiasis in recent years⁴³, with increased mortality rates⁴⁴, and echinocandins have been recommended for the treatment of invasive C. glabrata infections showing resistance to azoles. However, co-resistance to both echinocandins and azoles in clinical isolates of C. glabrata have been reported⁴⁵, with two cases of echinocandin-resistant C. glabrata infections recently reported from South Africa³⁶.

A new multi-drug resistant species, C. auris is taking the world by storm. First discovered in Japan⁴⁶, this species has been found in nine other countries on four continents. The Centre for Disease Control and Prevention (CDC) has issued a warning for increased awareness of C. auris in healthcare settings. This nosocomial pathogen is frequently misdiagnosed, shows resistance to different classes of antifungals routinely administered, and is associated with high mortality rates⁴⁷. The isolation of this species in South Africa⁴⁰ appears to be the only report in Africa at the time of writing this paper.

Regional differences in Candida susceptibility profiles have been observed. In South Africa, earlier reports of baseline data demonstrated a high susceptibility (100%) of C. albicans to fluconazole, along with intrinsically resistant non-albicans species^22,24, with more recent studies in South Africa showing an emerging resistance to azoles^23,25,48. The reasons for this change in susceptibility patterns is not clear, but it is worth noting that the earlier studies were done before the 2002 introduction of fluconazole as prophylaxis to patients attending HIV-AIDS clinics in South Africa⁴⁹.

Studies from abroad have reported cross-resistance to fluconazole in patients receiving itraconazole prophylaxis⁵⁰ and other previously administered azole therapies, such as ketoconazole and miconazole^51,52. Similar cross-resistance was recently reported in South Africa where 37% of C. parapsilosis isolates were susceptible to fluconazole and voriconazole, and 44% of fluconazole-resistant isolates were voriconazole cross-resistant⁴¹.

Studies from Bamenda²⁵ and Douala²⁶ in Cameroon showed high resistance of C. albicans isolates to azoles (>50% and 70% respectively), with low resistance reported from Mutegene²⁷ and Bamenda²⁵ for amphotericin B (4.9 and 4.3%) and 5-flucytosine (10.7 and 6.5%), respectively. The Douala study, on the other hand, reported increased C. albicans resistance to amphotericin B (52.6%) and 5-flucytosine (70%). A comparison of the Mutengene and Bamenda studies further revealed that C. dubliniensis and C. tropicalis susceptibilities differed between the two groups, with C. dubliniensis showing susceptibility to fluconazole and 100% resistance to amphotericin B in the Bamenda group, and increased resistance (66%) to fluconazole and no resistance to amphotericin B in the Mutengene group. C. tropicalis showed resistance to amphotericin B in the Bamenda group (50%) and only 4.3% resistance in the Mutegene group.

C. albicans resistance to amphotericin B has also been reported in Kenya (25.6%)³² and Ghana (23.4%)³⁹, while no resistance was seen in studies from Ivory Coast^30,38 nor Nigeria²⁹. Intermediate resistance values observed for clotrimazole and amphotericin B in studies from South West Cameroon, may indicate the need for administering higher doses to effectively treat these patients. This raises some concern, since both drugs are toxic at high concentrations and might have various side effects, such as the increase of blood pressure caused by clotrimazole therapy⁵³.

The application of topical antifungals, such as econazole and nystatin, is recommended for the localized treatment of Candida infections. Candida isolates from the Ivory Coast showed good susceptibility to nystatin with an increasing resistance noted in Ethiopia (1.3–4.7%), Kenya (36%), Gauteng, South Africa (67%) and Mutengene, Cameroon (68%). Resistance to econazole was reported in South West Cameroon^26,27. Overall, Candida isolates from Eastern African countries demonstrated the lowest resistance levels, with the exception of Kenya where resistance values for clotrimazole (74%) and nystatin (35.6%) were high³². Systemic antifungals are usually reserved for patients who are unresponsive to topical treatment in cases such as these.

The prevalence of Candida species isolated from sub-Saharan Africa appeared to follow the same trend as in other regions, with C. albicans being the predominant species, followed by C. glabrata. However, in a study from Cameroon²⁷ and two studies from Nigeria^28,29, C. tropicalis was the second most prevalent species. When comparing the drug resistance patterns with those in the rest of the world, C. albicans resistance to azoles was seen to be noticeably higher in Southwest South Africa²⁵ and two distinct regions in Cameroon^25,26, while C. glabrata resistance to azoles was found to be generally low. However, in a study from South Africa⁴¹, C. glabrata resistance to fluconazole was found to be similar to the high resistance levels seen in North America¹⁹. C. tropicalis resistance to fluconazole was seen to be noticeably high in East African countries, with up to 50% resistance seen in Tanzania³¹. Such high levels of resistance have not been documented in other regions.

Fluconazole is widely used in public health settings in the African continent and is used empirically in the treatment of systemic or localized Candida infections⁵⁴, as it is less toxic and regarded as more effective than imidazole antifungals, such as ketoconazole or amphotericin B, even though it is a teratogenic drug^55,56. Although still somewhat effective in other regions, the use of azoles as first line drugs for systemic infection should be revisited in certain areas of South and West Africa, due to their increasing inefficacy. Regular monitoring of Candida at a regional level could therefore be an important tool to aid in the prescription of antifungals based on the prevalent species and their susceptibilities to antifungal drugs in areas where routine microbiological laboratory testing is not available.

The sale of antimicrobial medications is largely unregulated in Africa and is exacerbated by the influx of fake and adulterated drugs with little or no active ingredients, often available both in pharmacies and on the streets. This problem is aggravated by practitioners who prescribe antimicrobial medications empirically based on clinical presentation, without prior knowledge of which microbial agent(s) are causing infections in their patients. These issues pose a serious public health threat, as they are largely responsible for more and more antimicrobial drugs being rendered ineffective in treating life-threatening infections. This is especially true in the case of the antifungal armamentarium, which is already very limited⁵⁷, especially in resource-poor settings. Limitations of this study include the paucity of available data from African regions, differences in sample sourcing, as well as techniques for isolation and susceptibility profiles from different regions, all of which complicate a comparison of outcomes of the cited studies.

The regional differences in antifungal drug susceptibility of Candida species, often seen within the same country, are an important finding that justifies the implementation of Candida species prevalence and susceptibility testing programmes in the African continent, notably in at-risk population groups, such as HIV patients. With the emergence of inherently drug resistant non-albicans species, more studies on Candida prevalence and drug susceptibility are needed throughout sub-Saharan Africa. This is most critical in resource-poor areas where there is little or no information available, such as southern (with the exception of South Africa) and central African countries and countries bordering the Sahara.

We would like to conclude by adding that Candida identification to species level is rarely made in clinical settings in Africa, and patients are treated empirically based on their clinical symptoms. The introduction of routine antimicrobial susceptibility testing before initiation of therapy can be relatively expensive, but is certainly a long-term cost effective solution in preventing the progression of drug resistance. Changes in drug susceptibility over time serve as a reminder for the need to test clinical Candida isolates for sensitivity to antifungal drugs in the effort to improve patient care and reduce patient morbidity and mortality.

Data availability

Dataset 1: Antifungal drug resistance of Candida species per region. DOI, 10.5256/f1000research.10327.d145319²¹